Corpora at NLP Centre

• ske.fi.muni.cz provides access to corpora at NLP Centre
• Corpus Architect – tool for building your own corpora, can be accessed after registration
• corpora (being) developed at NLP Centre:

  Corpus (lang)	M tokens
  Russian	20,162
  French	12,369
  Japanese	11,113
  American Spanish	8,719
  Arabic	6,637
  Czech	5,818
  Turkish	4,125
  czes (Czech)	465
  Kazakh	139
  Azerbaijani	115
  Tajik	52
  Uzbek	25
  Kyrgyz	24
  Turkmen	2
  DESAM (Czech)	1

• software for corpus processing:
  • NoSkech Engine – powerful and free corpus management system
  • chared – encoding detection
  • SpiderLing – intelligent web crawler
  • onion – de-duplication tool for text corpora
  • justext – boilerplate removing tool
• corpus related projects at NLP Centre:
  • Corpus Pattern Analysis
  • TIL corpus – corpus of formulas in Transparent Intensional Logic formulae
  • Kapradí

If necessary, contact corpadm@aurora.fi.muni.cz